Sweet/Fat Preference Taste in Subjects Who are Lean, Obese and Very Obese

Abstract

Purpose This study assessed the perception of sweetness, creaminess, and pleasantness from a sweet/fat preference test in subjects who are lean (BMI: 19–25), obese (BMI: 30–33) or very obese (BMI: 34–40) using categorical modeling. Methods Subjects tasted 16 dairy solutions consisting of 0%, 3.5%, 11.3% and 37.5% fat and each containing 0%, 5%, 10%, or 20% sugar and rated them for sweetness, creaminess and pleasantness. Results A proportional odds model described the perception of sweetness using an Emax for the effect of sugar and a linear effect for fat. Perception of creaminess was dependent on the fat and sugar content and was described with proportional odds model with linear effects of sugar and fat. Perception of pleasantness increased with sugar and fat but decreased in solutions containing 37.5% fat. A differential odds model using an Emax model for fat and sugar with a negative interaction between them allowed the sugar content to be less than proportional and the fat content to be greater than proportional for pleasantness. Conclusions Application of modeling provided understanding of these complex interactions of sugar and fat on the perception of sweetness, creaminess, and pleasantness and provides a tool to investigate obesity and pharmacological intervention.

Linking categorical models for prediction of pleasantness score using individual predictions of sweetness and creaminess: An advancement of categorical modeling

“…Data for modeling came from a single center study, approved by the Indiana University Institutional Review Board in 2006, and described in a previous paper by us [20]. All patients gave written informed consent and all procedures were performed in accordance with the ethical standards of the institutional research committee and Helsinki declaration.…”

“…Two sessions of scorings of all 16 solutions were performed and the solutions were presented in random order within each scoring session. The method for rating the solutions was previously described [20].…”

“…The previously described models were used to predict the individual responses for sweetness and creaminess using proportional odds models [20]. The details of the models are given in supplementary material S1.…”

“…Pleasantness has previously been described by a differential odds model which allowed for sugar to be less than proportional and amount of fat to be greater than proportional on the pleasantness score [20]. The effect of the amount of sugar and fat on pleasantness was described by Emax models with a negative apparent interaction between sugar and fat [20]. This model was used for comparison to the model developed replacing sugar and fat content with the individual prediction of sweetness and creaminess.…”

“…Recently, traditional PKPD models have been expanded to PK-PD-PD models to include the relationship of a PD response driving another PD response, which is a common biology of hormonal regulation [18,19]. The aim of this study was to develop linked categorical models, using the individual predictions of sweetness and creaminess from two separate categorical models as inputs into a differential odds model for the pleasantness response, and to compare the performance of this linked categorical model to a previously developed model which used the amount of sugar and fat in the solutions [20]. This was accomplished by using the same scoring data and assessing possible improvements between the models.…”